Hampton: Natural mortality and movement rates of Thunnus maccoyu 



607 



apparent fishing mortality inflicted upon the tagged 

 population by the Japanese fishery. Recall that one of 

 the assumptions of the HSH method is that fishing mor- 

 tality is maintained on the tagged population until com- 

 pletion of the experiment. However, as F approaches 

 zero for older age classes, overestimates of M result 

 from this model. To confirm this, the type 2 simula- 

 tions were rerun, with the only change being an in- 

 creased input value of q 3 such that F for the Japanese 

 fishery was increased to 0.2/year (the value derived 



from real data and previously used in the simulations 

 involving the HSH method was 0.007/year). Under 

 these conditions, the HSH model provided mean esti- 

 mates of M of 0.24/year for the analysis of experiment 

 2, experiment 3, and the two experiments pooled. This 

 agrees well with the value for M of 0.23/year input to 

 the simulation model. 



Discussion 



This paper represents the first specific 

 attempt to estimate the southern bluefin 

 tuna natural mortality rate from tagging 

 data. The simulation results confirm that 

 the method of Hearn et al. (1987) will 

 overestimate M substantially when the 

 apparent fishing mortality inflicted on 

 the tagged population in the latter part 

 of the experiment is very low. Also, the 

 sensitivity of the M estimates to the tag- 

 shedding model adopted makes interpre- 

 tation of the results difficult, particular- 

 ly as it was demonstrated by Hampton 

 and Kirkwood (1990) that long-term shed- 

 ding rates are very uncertain. 



